Adding upstream version 6.9.8.upstream/6.9.8

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

7 files changed, 77 insertions, 90 deletions

diff --git a/fs/bcachefs/bcachefs.h b/fs/bcachefs/bcachefs.h
index 91c3c1fef2..e6bcf74b0d 100644
--- a/fs/bcachefs/bcachefs.h
+++ b/fs/bcachefs/bcachefs.h
@@ -455,6 +455,7 @@ enum bch_time_stats {
 };
 
 #include "alloc_types.h"
+#include "btree_gc_types.h"
 #include "btree_types.h"
 #include "btree_node_scan_types.h"
 #include "btree_write_buffer_types.h"
@@ -485,49 +486,6 @@ enum bch_time_stats {
 
 struct btree;
 
-enum gc_phase {
-	GC_PHASE_NOT_RUNNING,
-	GC_PHASE_START,
-	GC_PHASE_SB,
-
-	GC_PHASE_BTREE_stripes,
-	GC_PHASE_BTREE_extents,
-	GC_PHASE_BTREE_inodes,
-	GC_PHASE_BTREE_dirents,
-	GC_PHASE_BTREE_xattrs,
-	GC_PHASE_BTREE_alloc,
-	GC_PHASE_BTREE_quotas,
-	GC_PHASE_BTREE_reflink,
-	GC_PHASE_BTREE_subvolumes,
-	GC_PHASE_BTREE_snapshots,
-	GC_PHASE_BTREE_lru,
-	GC_PHASE_BTREE_freespace,
-	GC_PHASE_BTREE_need_discard,
-	GC_PHASE_BTREE_backpointers,
-	GC_PHASE_BTREE_bucket_gens,
-	GC_PHASE_BTREE_snapshot_trees,
-	GC_PHASE_BTREE_deleted_inodes,
-	GC_PHASE_BTREE_logged_ops,
-	GC_PHASE_BTREE_rebalance_work,
-	GC_PHASE_BTREE_subvolume_children,
-
-	GC_PHASE_PENDING_DELETE,
-};
-
-struct gc_pos {
-	enum gc_phase		phase;
-	struct bpos		pos;
-	unsigned		level;
-};
-
-struct reflink_gc {
-	u64		offset;
-	u32		size;
-	u32		refcount;
-};
-
-typedef GENRADIX(struct reflink_gc) reflink_gc_table;
-
 struct io_count {
 	u64			sectors[2][BCH_DATA_NR];
 };
diff --git a/fs/bcachefs/btree_gc.c b/fs/bcachefs/btree_gc.c
index 791470b0c6..58f7c99e0e 100644
--- a/fs/bcachefs/btree_gc.c
+++ b/fs/bcachefs/btree_gc.c
@@ -1080,8 +1080,7 @@ fsck_err:
 
 static inline int btree_id_gc_phase_cmp(enum btree_id l, enum btree_id r)
 {
-	return  (int) btree_id_to_gc_phase(l) -
-		(int) btree_id_to_gc_phase(r);
+	return cmp_int(gc_btree_order(l), gc_btree_order(r));
 }
 
 static int bch2_gc_btrees(struct bch_fs *c, bool initial, bool metadata_only)
@@ -1126,7 +1125,7 @@ static void mark_metadata_sectors(struct bch_fs *c, struct bch_dev *ca,
 			min_t(u64, bucket_to_sector(ca, b + 1), end) - start;
 
 		bch2_mark_metadata_bucket(c, ca, b, type, sectors,
-					  gc_phase(GC_PHASE_SB), flags);
+					  gc_phase(GC_PHASE_sb), flags);
 		b++;
 		start += sectors;
 	} while (start < end);
@@ -1155,14 +1154,14 @@ static void bch2_mark_dev_superblock(struct bch_fs *c, struct bch_dev *ca,
 		b = ca->journal.buckets[i];
 		bch2_mark_metadata_bucket(c, ca, b, BCH_DATA_journal,
 					  ca->mi.bucket_size,
-					  gc_phase(GC_PHASE_SB), flags);
+					  gc_phase(GC_PHASE_sb), flags);
 	}
 }
 
 static void bch2_mark_superblocks(struct bch_fs *c)
 {
 	mutex_lock(&c->sb_lock);
-	gc_pos_set(c, gc_phase(GC_PHASE_SB));
+	gc_pos_set(c, gc_phase(GC_PHASE_sb));
 
 	for_each_online_member(c, ca)
 		bch2_mark_dev_superblock(c, ca, BTREE_TRIGGER_GC);
@@ -1773,7 +1772,7 @@ int bch2_gc(struct bch_fs *c, bool initial, bool metadata_only)
 	if (ret)
 		goto out;
 again:
-	gc_pos_set(c, gc_phase(GC_PHASE_START));
+	gc_pos_set(c, gc_phase(GC_PHASE_start));
 
 	bch2_mark_superblocks(c);
 
@@ -1800,7 +1799,7 @@ again:
 		 */
 		bch_info(c, "Second GC pass needed, restarting:");
 		clear_bit(BCH_FS_need_another_gc, &c->flags);
-		__gc_pos_set(c, gc_phase(GC_PHASE_NOT_RUNNING));
+		__gc_pos_set(c, gc_phase(GC_PHASE_not_running));
 
 		bch2_gc_stripes_reset(c, metadata_only);
 		bch2_gc_alloc_reset(c, metadata_only);
@@ -1827,7 +1826,7 @@ out:
 
 	percpu_down_write(&c->mark_lock);
 	/* Indicates that gc is no longer in progress: */
-	__gc_pos_set(c, gc_phase(GC_PHASE_NOT_RUNNING));
+	__gc_pos_set(c, gc_phase(GC_PHASE_not_running));
 
 	bch2_gc_free(c);
 	percpu_up_write(&c->mark_lock);
diff --git a/fs/bcachefs/btree_gc.h b/fs/bcachefs/btree_gc.h
index 607575f83a..5c97eb6a38 100644
--- a/fs/bcachefs/btree_gc.h
+++ b/fs/bcachefs/btree_gc.h
@@ -3,6 +3,7 @@
 #define _BCACHEFS_BTREE_GC_H
 
 #include "bkey.h"
+#include "btree_gc_types.h"
 #include "btree_types.h"
 
 int bch2_check_topology(struct bch_fs *);
@@ -35,38 +36,17 @@ int bch2_gc_thread_start(struct bch_fs *);
 /* Position of (the start of) a gc phase: */
 static inline struct gc_pos gc_phase(enum gc_phase phase)
 {
-	return (struct gc_pos) {
-		.phase	= phase,
-		.pos	= POS_MIN,
-		.level	= 0,
-	};
-}
-
-static inline int gc_pos_cmp(struct gc_pos l, struct gc_pos r)
-{
-	return  cmp_int(l.phase, r.phase) ?:
-		bpos_cmp(l.pos, r.pos) ?:
-		cmp_int(l.level, r.level);
-}
-
-static inline enum gc_phase btree_id_to_gc_phase(enum btree_id id)
-{
-	switch (id) {
-#define x(name, v, ...) case BTREE_ID_##name: return GC_PHASE_BTREE_##name;
-	BCH_BTREE_IDS()
-#undef x
-	default:
-		BUG();
-	}
+	return (struct gc_pos) { .phase	= phase, };
 }
 
-static inline struct gc_pos gc_pos_btree(enum btree_id id,
+static inline struct gc_pos gc_pos_btree(enum btree_id btree,
 					 struct bpos pos, unsigned level)
 {
 	return (struct gc_pos) {
-		.phase	= btree_id_to_gc_phase(id),
-		.pos	= pos,
+		.phase	= GC_PHASE_btree,
+		.btree	= btree,
 		.level	= level,
+		.pos	= pos,
 	};
 }
 
@@ -91,6 +71,22 @@ static inline struct gc_pos gc_pos_btree_root(enum btree_id id)
 	return gc_pos_btree(id, SPOS_MAX, BTREE_MAX_DEPTH);
 }
 
+static inline int gc_btree_order(enum btree_id btree)
+{
+	if (btree == BTREE_ID_stripes)
+		return -1;
+	return btree;
+}
+
+static inline int gc_pos_cmp(struct gc_pos l, struct gc_pos r)
+{
+	return   cmp_int(l.phase, r.phase) ?:
+		 cmp_int(gc_btree_order(l.btree),
+			 gc_btree_order(r.btree)) ?:
+		-cmp_int(l.level, r.level) ?:
+		 bpos_cmp(l.pos, r.pos);
+}
+
 static inline bool gc_visited(struct bch_fs *c, struct gc_pos pos)
 {
 	unsigned seq;
diff --git a/fs/bcachefs/btree_gc_types.h b/fs/bcachefs/btree_gc_types.h
new file mode 100644
index 0000000000..b82c24bcc0
--- /dev/null
+++ b/fs/bcachefs/btree_gc_types.h
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _BCACHEFS_BTREE_GC_TYPES_H
+#define _BCACHEFS_BTREE_GC_TYPES_H
+
+#include <linux/generic-radix-tree.h>
+
+enum gc_phase {
+	GC_PHASE_not_running,
+	GC_PHASE_start,
+	GC_PHASE_sb,
+	GC_PHASE_btree,
+};
+
+struct gc_pos {
+	enum gc_phase		phase:8;
+	enum btree_id		btree:8;
+	u16			level;
+	struct bpos		pos;
+};
+
+struct reflink_gc {
+	u64		offset;
+	u32		size;
+	u32		refcount;
+};
+
+typedef GENRADIX(struct reflink_gc) reflink_gc_table;
+
+#endif /* _BCACHEFS_BTREE_GC_TYPES_H */
diff --git a/fs/bcachefs/ec.c b/fs/bcachefs/ec.c
index 556a217108..0a49c2e995 100644
--- a/fs/bcachefs/ec.c
+++ b/fs/bcachefs/ec.c
@@ -880,7 +880,7 @@ static int __ec_stripe_mem_alloc(struct bch_fs *c, size_t idx, gfp_t gfp)
 	if (!genradix_ptr_alloc(&c->stripes, idx, gfp))
 		return -BCH_ERR_ENOMEM_ec_stripe_mem_alloc;
 
-	if (c->gc_pos.phase != GC_PHASE_NOT_RUNNING &&
+	if (c->gc_pos.phase != GC_PHASE_not_running &&
 	    !genradix_ptr_alloc(&c->gc_stripes, idx, gfp))
 		return -BCH_ERR_ENOMEM_ec_stripe_mem_alloc;
 
diff --git a/fs/bcachefs/sb-downgrade.c b/fs/bcachefs/sb-downgrade.c
index a98ef940b7..ed6d298bc1 100644
--- a/fs/bcachefs/sb-downgrade.c
+++ b/fs/bcachefs/sb-downgrade.c
@@ -134,7 +134,8 @@ downgrade_entry_next_c(const struct bch_sb_field_downgrade_entry *e)
 #define for_each_downgrade_entry(_d, _i)						\
 	for (const struct bch_sb_field_downgrade_entry *_i = (_d)->entries;		\
 	     (void *) _i	< vstruct_end(&(_d)->field) &&				\
-	     (void *) &_i->errors[0] < vstruct_end(&(_d)->field);			\
+	     (void *) &_i->errors[0] <= vstruct_end(&(_d)->field) &&			\
+	     (void *) downgrade_entry_next_c(_i) <= vstruct_end(&(_d)->field);		\
 	     _i = downgrade_entry_next_c(_i))
 
 static int bch2_sb_downgrade_validate(struct bch_sb *sb, struct bch_sb_field *f,
@@ -142,7 +143,17 @@ static int bch2_sb_downgrade_validate(struct bch_sb *sb, struct bch_sb_field *f,
 {
 	struct bch_sb_field_downgrade *e = field_to_type(f, downgrade);
 
-	for_each_downgrade_entry(e, i) {
+	for (const struct bch_sb_field_downgrade_entry *i = e->entries;
+	     (void *) i	< vstruct_end(&e->field);
+	     i = downgrade_entry_next_c(i)) {
+		/*
+		 * Careful: sb_field_downgrade_entry is only 2 byte aligned, but
+		 * section sizes are 8 byte aligned - an empty entry spanning
+		 * the end of the section is allowed (and ignored):
+		 */
+		if ((void *) &i->errors[0] > vstruct_end(&e->field))
+			break;
+
 		if (BCH_VERSION_MAJOR(le16_to_cpu(i->version)) !=
 		    BCH_VERSION_MAJOR(le16_to_cpu(sb->version))) {
 			prt_printf(err, "downgrade entry with mismatched major version (%u != %u)",
@@ -214,7 +225,7 @@ int bch2_sb_downgrade_update(struct bch_fs *c)
 
 		dst = (void *) &darray_top(table);
 		dst->version = cpu_to_le16(src->version);
-		dst->recovery_passes[0]	= cpu_to_le64(src->recovery_passes);
+		dst->recovery_passes[0]	= cpu_to_le64(bch2_recovery_passes_to_stable(src->recovery_passes));
 		dst->recovery_passes[1]	= 0;
 		dst->nr_errors		= cpu_to_le16(src->nr_errors);
 		for (unsigned i = 0; i < src->nr_errors; i++)
diff --git a/fs/bcachefs/super-io.c b/fs/bcachefs/super-io.c
index bfdb15e7d7..a337246e91 100644
--- a/fs/bcachefs/super-io.c
+++ b/fs/bcachefs/super-io.c
@@ -1123,18 +1123,12 @@ bool bch2_check_version_downgrade(struct bch_fs *c)
 	 * c->sb will be checked before we write the superblock, so update it as
 	 * well:
 	 */
-	if (BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb) > bcachefs_metadata_version_current) {
+	if (BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb) > bcachefs_metadata_version_current)
 		SET_BCH_SB_VERSION_UPGRADE_COMPLETE(c->disk_sb.sb, bcachefs_metadata_version_current);
-		c->sb.version_upgrade_complete = bcachefs_metadata_version_current;
-	}
-	if (c->sb.version > bcachefs_metadata_version_current) {
+	if (c->sb.version > bcachefs_metadata_version_current)
 		c->disk_sb.sb->version = cpu_to_le16(bcachefs_metadata_version_current);
-		c->sb.version = bcachefs_metadata_version_current;
-	}
-	if (c->sb.version_min > bcachefs_metadata_version_current) {
+	if (c->sb.version_min > bcachefs_metadata_version_current)
 		c->disk_sb.sb->version_min = cpu_to_le16(bcachefs_metadata_version_current);
-		c->sb.version_min = bcachefs_metadata_version_current;
-	}
 	c->disk_sb.sb->compat[0] &= cpu_to_le64((1ULL << BCH_COMPAT_NR) - 1);
 	return ret;
 }